MuXodious/LFM2-8B-A1B-absolute-heresy
Abliterated vs LiquidAI/LFM2-8B-A1B. Looks like abliteration (refusal removal): only the output projections (out_proj, w2) changed, in 11 of 24 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: MLP 192/2162, Embeddings 0/2, other 3/54, Norms 0/60, Attention 4/24
3 derivatives checked
- mradermacher/LFM2-8B-A1B-absolute-heresy-i1-GGUF: Quantization
- mradermacher/LFM2-8B-A1B-absolute-heresy-GGUF: Same weights, converted
- MuXodious/LFM2-8B-A1B-absolute-heresy-GGUF: Quantization